Direct support for hypotheses is finding its way to high-profile journals

By | April 23, 2013

In this week's Nature paper, Tingley & Huybers report that recent temperature extremes at high northern latitudes are unprecedented in the past 600 years. Besides the scientific relevance (which I do not discuss here) it has one remarkable methodological aspect: it uses hierarchical Bayesian modelling. Moreover, what is really exciting is the way the authors report their results in the very abstract of the paper. They say:

"The summers of 2005, 2007, 2010 and 2011 were warmer than those of all prior years back to 1400 (probability P > 0.95), in terms of the spatial average. The summer of 2010 was the warmest in the previous 600 years in western Russia (P > 0.99) and probably the warmest in western Greenland and the Canadian Arctic as well (P > 0.90)."

Can you see how different this is from all of the other papers that have been published during the last century or so? The "P values" are sort of reversed! There is no P<0.001. The authors are not rejecting hypotheses any more. They are reporting direct support for them! The authors are saying how probable it is that something happened, not how probable it is to observe an arrangement of data under an assumption of a null hypothesis.

We are witnessing a paradigm shift in scientific thinking. This is it, it is happening now. The fact that scientists are reporting direct support for hypotheses in a Nature paper indicates that the crack in the dam has already turned into a massive stream. The structure that had supported the Frequentist paradigm does not hold any more. I feel like celebrating.

Tingley M.P. & Huybers P. (2013) Recent temperature extremes at high northern latitudes unprecedented in the past 600 years. Nature, 496, 201-205.

Leave a Reply

Your email address will not be published. Required fields are marked *